What is the oral bioavailability of J147?

Approximately 28 per cent, measured in mice after a single 20 mg/kg dose and reported alongside the aged-mouse efficacy work in Alzheimer's Research and Therapy in 2013.

What is the half-life of J147?

About 1.5 hours in plasma and about 2.5 hours in brain, in mice. The brain-to-blood ratio is around 0.5, so the compound crosses into the brain and does not accumulate there for long.

Pharmacokinetics · Dosage form · Capability

Twenty-eight per cent, and ninety minutes

Q: Why does J147 need formulation work?

Because a compound that is 28 per cent orally available with a 90-minute plasma half-life has its exposure decided by its dosage form. The only human study of J147 used a corn-oil vehicle. Where a molecule's active window is short, the presentation determines how much of the pharmacology is ever expressed.

J147's published mouse pharmacokinetics are approximately 28 % oral bioavailability, a 1.5-hour plasma half-life, a 2.5-hour brain half-life and a brain-to-blood ratio near 0.5. A compound with that profile has its exposure decided by its dosage form. The only human study of J147 used a corn-oil vehicle.

These numbers are not a criticism of the molecule. They are a specification for the thing that has to be built around it.

28 %oral bioavailability — 72 % of an oral dose never becomes exposure

1.5 hplasma half-life — the active window is short and the schedule matters

2.5 hbrain half-life — longer than plasma, which is the useful asymmetry

0.5brain-to-blood ratio — it crosses readily; it does not accumulate

Single 20 mg/kg dose in mice; Prior et al., Alzheimer's Research & Therapy, 2013. Mouse pharmacokinetics do not transfer directly to any other species, and are used here as a specification for formulation work, not as a human prediction.

Schematic, not data: the shape of the problem a short half-life and modest oral availability create. The curve is drawn rather than plotted, and the threshold line is notional — no efficacious human concentration for J147 has been published. Original diagram, Panacea Bio Chem.

What a profile like this asks of a dosage form

Three things, none of them exotic, all of them engineering:

1 · Get more of the dose in

At 28 % oral availability the largest single lever is absorption. That is solubility, dispersion state, and the vehicle — the reason the Phase 1's corn-oil vehicle is worth noticing rather than passing over.

2 · Keep the material intact until it is used

Stability of the solid, stability in solution, and behaviour under freeze-drying. A compound is only as good as what is left of it at the point of dosing, and that is an analytical question before it is a formulation one.

3 · Control the moment of reconstitution

With a short half-life, when a solution is made and how quickly it is used stop being housekeeping and start being variables. A presentation that reconstitutes reproducibly at the point of use removes one of them.

4 · Know the exposure, rather than assume it

So that when an experiment gives a negative result, the negative belongs to the hypothesis and not to a vial that had less in it than the label said.

The Panacea Bio Chem capability

This is where a molecule like J147 becomes usable or stays a curiosity, and it is the work Panacea Bio Chem exists to do. Every technology below was invented by Bogdan Dicoias and built in-house, because the compound's exposure is decided by equipment somebody has to own. The J147 programme is ongoing and directional; no result is claimed for it here.

Capability	What it is	How it bears on J147
SYNTHESERACT	Our synthesis platform and the cleanest route we have to material we control end to end.	Identity, purity and impurity profile established in-house rather than inherited from a certificate somebody else wrote.
Lyochrysalis™	Our lyophilisation machine — the system the cake is dried in.	Solid-state stability is decided during drying. What the cake becomes here is what is dosed months later.
LyoLevit™	Zero-contact orbital sublimation inside Lyochrysalis: the cake levitates and spins rather than sitting on a shelf, so heat arrives as 360° radiation instead of floor conduction.	The sublimation surface goes from the top face alone to very nearly the whole cake, and nothing adheres to a wall. Drying that is faster and more even is drying that damages less of what you made.
Lyoprester™	The product cartridge — a two-chamber pre-filled syringe. Chamber 1 holds the lyophilised cake under vacuum with argon in a trehalose glass; Chamber 2 holds the P-EARL liquid, bubble-free.	Tapping the plunger lets the Chamber-1 vacuum draw the liquid through a channel into the cake: reconstitution happens in situ, with no separate mixing step and no open vial. For a compound with a 1.5-hour half-life, that removes the largest uncontrolled variable between the vial and the animal.
P-EARL™	Panacea-Engineered Aseptic Reconstitution Liquid: the Chamber-2 preload, a buffer / anti-gelation / chelation stack tuned per compound.	The liquid a compound is reconstituted into is part of its formulation, not a diluent. Tuning it is how a redissolved cake behaves like the material that went in.
OxyDeplete™ / ArgonLock™	Degassing plus a no-headspace doctrine — the oxygen-starved seal — finished with a finalinert-atmosphere lock under argon.	Relevant to any compound whose degradation route runs through oxygen, and to any shelf-life claim that has to survive two years rather than a week.
Dicoias Ψ™	Our formulation mathematics: every substance and every excipient reduced to a vector on the same six axes, with a Ψ operator family that scores the recipe and prescribes the additions. Set out in full below.	It is the layer that decides what a J147 cartridge contains, and it decides it as an explicit computation with a stated shortfall rather than as a preference.
Peptourbillon™	The formulation calculator: layer masses, excipient ratios, volumes and reconstitution parameters for the cartridges.	It turns the mathematics into the numbers a run actually needs.
S3Pulse™	The control system — the brain that runs Lyochrysalis, holds the process doctrine and keeps the run record.	The place a batch stops being an intention and becomes a recorded derivation.

Research use only

J147 is an investigational compound. It is not approved as a medicine in any territory, it is not supplied here for human use, and nothing on this page is dosing advice. Nothing here is medical advice.

Next: chemistry and freedom to operate Published synthesis routes, analytical characterisation, and an honest account of the patent position — including what we have not established.

The mathematics: Dicoias Ψ™

Formulation is usually argued. At Panacea Bio Chem it is computed. Dicoias Ψ™ is our own formulation mathematics, and its premise is that mathematics is the universal formulation language: every substance is reduced to a vector, every excipient is reduced to a vector in the same space, and a Ψ operator family scores the recipe and prescribes what to add. It is named for its author, and it is the layer that decides what goes into a cartridge before anything is weighed.

P →(H, O) N →(Λ) Ψ →(τ) Score

That line is the whole pipeline, and each arrow is a defined transform rather than a judgement call.

P — the intrinsic signature

Four axes derived from the molecule alone, with no formulation context allowed in: D1 electrostatic sensitivity, D2 hydrophobic and interfacial burden, D3 chemical liability, D4 solid-state matrix dependence — seventeen sub-components beneath them. This is what the substance is, before anyone decides what to do with it.

H and O — context, through eight operators

The context vector H carries pH, ionic strength, concentration, storage and process temperature, oxygen and light exposure, interfacial stress, freeze-concentration severity and drying profile; a second pass adds what is only knowable once a candidate exists — T_g gap, residual moisture, matrix state, collapse margin. Context is applied to sub-components first and aggregated second, through a closed registry of eight approved operators — direct evaluation, Gaussian proximity kernel, linear amplifier, saturating screening, sigmoid, threshold gate, Henderson—Hasselbalch ionisation, Arrhenius/Eyring acceleration. Forty-two edges, each one an instance with its own grounding, validity domain, units and confidence. No coefficient enters this system without a name and a provenance.

# context couples to sub-components, then aggregates n_jk = d_jk × g_jk(H) context-coupled transform N_j = Σ_k w_jk × n_jk axis-level protective demand

N — six axes of protective demand

The need vector is deliberately wider than the signature that produced it, because one intrinsic property can create two unlike demands:

N1aElectrostatic proximity — the pI window

N1bNet-charge burden at the formulation pH

N2Aggregation and interfacial stress

N3Chemical and oxidative liability

N4Solid-state / matrix dependence

N5Molecular crowding and viscosity

Those labels are not our prose for this page; they are the strings the running system displays, because a number is never shown here without its meaning. And the semantics are fixed: N is protective demand under a stated context — not risk, not probability, not a score. Excipients satisfy needs.

Λ — excipients in the same coordinates

Every additive carries a capability vector on those same six axes, signed, so that an excipient which helps one axis and harms another says so in the arithmetic rather than in a footnote. A hard-constraint gate runs first: an excipient that fails its constraints is rejected and its capability is never computed at all.

Λ_x,eff = m_x ⊙ c_x ⊙ (Λ⁺− Λ⁻) ∈ [−1, 1]⁶ signed effective capability C = clip_[−1,1]( Σ_{x ∈ S} Λ_x,eff ) capability of the whole mixture

Ψ — the compatibility potential

Here is the part that matters, and the part most scoring systems get wrong. Ψ is not a reward dot product. It is a shortfall energy, and lower is better: it measures how badly a candidate set fails to cover what the substance actually needs.

# gate before score: an infeasible formulation gets no partial credit if any Γ_x fails: Ψ(S | P, H) = +∞ # otherwise, per axis j: C⁺_j = max(C_j, 0) positive coverage C⁻_j = max(−C_j, 0) active harm Δ_j = max(0, N_j − C⁺_j) unmet need E_j = max(0, C⁺_j − N_j) oversupply Ψ(S | P, H) = Σ_j=1⁶ π_j [ α×Δ_j² + β×(C⁻_j)² + γ×E_j² ] # v1 weights α = 1.00 shortfall — the primary driver β = 2.00 harm — penalised twice as hard as failing to help γ = 0.15 oversupply — mild, so safety margins stay allowed

Two design decisions inside that expression carry the doctrine. β > α is first, do no harm written as arithmetic: actively worsening a formulation costs twice what passively failing to protect it costs, and it costs that on every axis, including the ones this particular substance barely needs. And the terms are squared, which makes the objective non-compensatory — one large uncovered gap is worse than two moderate ones, so a formulation cannot buy its way out of a serious deficit with a pile of irrelevant coverage. Cost, complexity and preference are deliberately not in Ψ; they are tie-breakers, never terms.

Score, and the ranking that is not a weighted sum

Dicoias Score = 100 × exp(−Ψ / τ) Ψ=0 → 100; Ψ→∞ → 0 # candidates rank lexicographically, never by adding unlike things together 1. lowest Ψ compatibility energy 2. fewest excipients Occam 3. highest mean confidence epistemic preference

The honesty mechanism is part of the mathematics rather than a disclaimer bolted to it: coverage is accounted explicitly, and certainty is capped for substances the system has not seen enough of. An unfamiliar molecule does not get a confident number just because the arithmetic will happily produce one.

What this has to do with J147

J147 is a small molecule, not a peptide, and we will say plainly which parts of the space carry over and which do not. The axes anchored on isoelectric point — N1a and N1b — are peptide-anchored and do not transfer as written. N3 does: chemical and oxidative liability is exactly the axis a hydrazide with an electron-rich methoxy-phenyl ring loads. N4 does: a compound presented as a lyophilised cake lives or dies on solid-state matrix dependence. N2 and N5 apply to any poorly water-soluble compound asked to redissolve completely in a small volume. Extending the Ψ space beyond peptides — a formulate-anything engine — is active work here, and J147 is one of the molecules driving it.